Researchers Secure Quantum Computation on Untrusted Hardware with New Encryption Framework

Researchers at University of the Basque Country, led by Jon Hernández-Bueno, have developed a new universal quantum homomorphic encryption framework maintaining information-theoretic security. This framework, derived from the Quantum One-Time Pad scheme, supports a wide class of quantum operations on encrypted quantum states and enables non-interactive homomorphic evaluation of circuits expressible in the Clifford’T gate set. This represents a key advance towards practical and private computation on near-term quantum hardware. Experimental validation using both simulated and real quantum processors confirms the correctness of the homomorphic operations and preservation of key secrecy despite circuit noise and device constraints. Quantum One-Time Pad enables secure computation on encrypted quantum states A quantum homomorphic encryption framework, termed QOTPH, now enables computations on encrypted quantum data, bridging a longstanding gap between theoretical proposals and practical application. Fully homomorphic and practically implementable quantum encryption has previously remained elusive due to the inherent challenges of manipulating quantum information without disturbing its superposition and entanglement. This new approach supports a broad class of quantum operations on encrypted quantum states utilising the Clifford’T gate set, a foundational step towards secure quantum computation on untrusted hardware. The significance lies in the potential to outsource quantum computations to third-party quantum processors without compromising the confidentiality of the input data or the algorithm itself. Exploiting the Quantum One-Time Pad scheme and Pauli encryption, QOTPH enables non-interactive homomorphic evaluation, maintaining information-theoretic security, a level of protection not previously achievable with existing methods. Information-theoretic security implies that the encryption scheme is secure regardless of the computational power of the adversary, unlike computational security which relies on the presumed difficulty of certain mathematical problems.

The Quantum One-Time Pad, a symmetric encryption method, requires a key as long as the message being encrypted, ensuring perfect secrecy if the key is used only once. In the quantum realm, this translates to a random quantum state being used to encrypt each qubit of the input data. Pauli encryption, a specific encoding scheme, is employed to map quantum information onto a Pauli basis, facilitating the homomorphic operations. Even when circuits introduce noise, a pervasive issue in current quantum devices, the correctness of homomorphic operations and key secrecy was validated. This resilience to noise is crucial for practical implementation on real quantum processors. The QOTPH framework successfully encrypted quantum states and performed computations using this standard for quantum algorithms, allowing for both arbitrary circuits and controlled, parameterised operations important for advanced techniques like variational quantum algorithms. Variational quantum algorithms, used in areas like quantum chemistry and machine learning, require repeated evaluation of quantum circuits with varying parameters, making homomorphic encryption particularly valuable for preserving privacy during optimisation processes. Both simulated environments and actual quantum processors were used to test the system, confirming its functionality across different platforms. Although initial demonstrations were limited in scale, extending it to handle larger and more complex quantum circuits remains a substantial hurdle to widespread practical application; data could be processed without constant communication between parties, enhancing efficiency and security. The current implementation demonstrates the feasibility of processing up to a limited number of qubits, with future work focused on scaling the scheme to accommodate larger quantum registers. Scalable quantum data processing secured via a restricted gate set Intense interest in quantum homomorphic encryption is being driven by the promise of confidential computation, allowing processing of data without revealing its contents. This is particularly relevant in scenarios such as secure quantum machine learning, confidential quantum key distribution, and privacy-preserving quantum simulations. However, this new framework currently relies on the Clifford+T gate set, a restricted set of tools for quantum operations. The Clifford’T gate set comprises the Clifford gates, which can be efficiently simulated classically, and the non-Clifford T gate, essential for achieving quantum speedup. While limiting the universality of the scheme, this restriction allows for a more manageable implementation and potential for future expansion, despite presenting a hurdle for truly universal quantum computation. The choice of the Clifford+T gate set is a pragmatic one, balancing expressiveness with the complexity of implementing homomorphic operations. Implementing homomorphic operations for arbitrary quantum gates is significantly more challenging and resource-intensive. Built upon the Quantum One-Time Pad, this symmetric encryption method offers a scalable route to privacy-preserving computation on emerging quantum hardware, paving the way for more complex schemes. Symmetric encryption, where the same key is used for both encryption and decryption, is generally faster and more efficient than asymmetric encryption, making it suitable for resource-constrained quantum devices. QOTPH, utilising a random key as long as the data itself, allows computations on encrypted quantum information without needing to decrypt it first. This is achieved by carefully designing the homomorphic operations such that they preserve the encryption structure, ensuring that the decrypted output is equivalent to the result of the computation performed on the original plaintext. This approach offers a scalable solution for protecting privacy when using emerging quantum hardware and represents a vital step towards practical, confidential data processing. The framework’s scalability is further enhanced by its ability to process quantum data in a parallel manner, leveraging the inherent parallelism of quantum computation. Future research will focus on exploring techniques to reduce the key size and improve the efficiency of the homomorphic operations, further enhancing the scalability and practicality of the QOTPH framework. The development of more efficient quantum error correction codes will also be crucial for mitigating the effects of noise and enabling reliable computation on larger quantum circuits. The researchers demonstrated a new framework, QOTPH, for performing secure quantum computation on encrypted data. This method maintains information-theoretic security by building upon the Quantum One-Time Pad, enabling computations without decrypting the quantum information first. QOTPH supports operations within the Clifford+T gate set and is validated through simulations and experiments on real quantum processors. The framework offers a scalable solution for privacy-preserving computation on quantum hardware, and future work aims to reduce key size and improve operational efficiency. 👉 More information🗞 Quantum Homomorphic Encryption: Towards Practical and Private Computation on Untrusted Quantum Hardware🧠 ArXiv: https://arxiv.org/abs/2604.19256 Tags: